What serverless means and why it is cheaper for spiky traffic

The short answer

Serverless is a way of running code where you never manage a server. You write small functions, a form handler, a search endpoint, a webhook receiver, and the provider runs each one on demand, on its own machines, scaling up when many requests arrive and down to zero when none do. You are billed per call and per millisecond of running time. Nothing runs, and nothing costs, when nobody is using it.

That last property is why it fits business websites so well: their traffic is uneven, and idle time is most of the time.

What changes compared with a server

	A server you rent	Serverless functions
What you manage	Operating system, web server, runtime, patching, capacity	Your function’s code and its dependencies
When it runs	Always, whether busy or idle	Only when called
What you pay for	The machine, per month, at the size you chose	Calls and running time
Traffic spike	Falls over, or you over-provisioned all year for this moment	Scales automatically
Quiet night	Costs the same as a busy day	Costs nothing
Security patching of the platform	Yours	The provider’s
Where it runs	One location	Often close to the visitor

Why it is cheaper for spiky traffic

A business website is busy for minutes and idle for hours. A newsletter goes out and a thousand people arrive at once; the rest of the day, a handful. A server has to be sized for the thousand and paid for during the handful. Serverless is sized for nothing and bills for the thousand calls, then goes quiet. The idle time, which is most of the time, costs zero.

The trade-offs

Where it fits on a business site

The pages are static: built once, served from the edge, no function needed. The interactive parts that genuinely need a server, a contact form that must validate and send, a search endpoint, a payment webhook, a small integration with your CRM, are functions. Each runs for milliseconds when used and costs nothing otherwise. That combination is our default hosting for business websites, and serverless is what makes the dynamic parts cheap and maintenance-free.

What this means for you

You do not need to know how functions work. You need to know that the parts of your site that occasionally need to do something can run without a server, scale on their own, and cost nothing at rest. If your site still runs on a machine that is always on, mostly idle, and patched by somebody you hope remembers, serverless next to a static site removes the machine, the idle cost and the patching in one move.